Research on AI-driven complex network and management system of coal and gas outburst accident

Abstract

Coal and gas outburst accidents are the most dangerous and complex accidents in coal mines. There is an urgent need for a systematic, objective, intelligent, and full process coal and gas outburst accident control system. The study constructs artificial intelligence (AI) -driven complex network and control system for coal and gas outburst accident: a coal and gas outburst accident control method based on text mining, complex network, and knowledge graph is proposed to achieve systematic extraction of causal factors, accurate identification of control strategy, intelligent formulation of control measures for coal and gas outburst accidents. The research results indicate that: (1) 70 causal factors are extracted from 72 investigation reports on coal and gas outburst accidents, including 21 human factors, 16 mechanical equipment factors, 14 environmental factors, and 19 management factors. (2) The coal and gas outburst accident complex network has a more complex structure, stronger robustness, and more prominent multi factor coupling effect, requiring more intelligent control methods. (3) The attack strategy generated by the innovatively proposed comprehensive value (CV) topology indicator is the optimal control strategy for coal and gas outburst accidents, maintaining higher attack efficiency in multiple stages and better controlling accident complex network. It also performs well in controlling single accident. (4) The coal and gas outburst accident control technology which is based on knowledge graph can quickly match the control measures and relationships, perform intelligent search of entities, reduce the time and energy consumption of managers for knowledge analysis, and provide intelligent decision support for managers. This study validates the scientific and practical feasibility of AI in the control of coal and gas outburst accidents. Future research will focus on further expanding data sources, clarifying the mechanism of coal and gas outburst accidents, and enhancing the application of expert systems, with the goal of more comprehensively improving coal mine accident control capabilities.